How does it work?

The simplified model of software working  in the system is shown below. The model is shown using UML 2.0 Activity diagram notation.

It is stream model. Here are some comments to the  steps of the model below:

  1. Real web camera in the system is sending video stream down the graph frame by frame.
  2. software takes each frame and clear it from noise
  3. Video frame is passed to the next module, which normalise the brightness and white balance. Modern cameras and camera soft is trying to be smart and to adjust lightness of the picture according to the actual light, but the side effect of this is that every move can change the lightness and color of the hole scene different from previous frame. It is impossible to detect figure correctly if the hole picture is changing. So this module remember previous frames and compensate small changes of light and color of the hole video frame.
  4. On this step detects where is the figure and what is background on the video. The principle of work is very complex and need more then one article to describe.
    As the result of work  the figure mask is created, it shows the area of the picture on which the figure was detected.
  5. On the next step animation frame (snapshot of flash movie is overlain the video frame, but only those areas which are detected as background according to figure mask
  6. Pretty the same as previous step but animation is applied over figure
  7. And finally the foreground animation is placed over the hole video frame. Please keep in mind that flash animation has alpha channel (some areas could be transparent).
    The video is ready
  8. It is passed to the emulator of webcam - Virtual camera
  9. Video chat or video call software, like Skype, Camfrog, Paltalk, SightSpeed could  use virtual camera as  real one and broadcast video with flash animated effects to your friends or chat community.

How webcam effects work

Enter Your Account

Forget your password?